Using AI to Win in Business

Understanding AI's Role in Data-Driven Business Models: Unpacking the strategic rationale for embedding AI into existing enterprises and examining its influence on wider market trends.

As I continue my travels this week, I have delved into the strategic implications of Artificial Intelligence (AI) for businesses. In this post, I present insights and analyses that go beyond theoretical applications, focusing on practical, strategic considerations that businesses are likely to incorporate into their AI adoption strategies. These insights are framed not only to aid individual businesses in understanding and leveraging AI effectively but also to provide a broader perspective on where value can be captured within the AI and data analysis value chain on an industry-wide scale.

---

Frontier Models are Basic Platforms

First off, it is worth putting frontier models in context for business. In the rapidly evolving landscape of Artificial Intelligence (AI), frontier models are in a category of their own. While numerous powerful AI models have become accessible through open-source platforms, it's the frontier models that are set to redefine the AI paradigm. These advanced models are not just iterations of existing technology; they represent a leap into a future dominated by a few global entities capable of sustaining the massive capital and resource investments required for their development.

The Race for Scale

Unprecedented Scale and Complexity: Frontier models are distinguished by their sheer scale and complexity, which demand vast computational resources and sophisticated development processes. This scale isn't just about size; it's about the ability to process and learn from an exponentially larger set of data, leading to more nuanced and accurate AI capabilities.

An Arms Race for AI Dominance: As outlined in the article "AGI: Scale Is All You Need," the development of frontier models is akin to an arms race, with scale being the critical factor. Model size has increased by 17.6x each year for the last decade. Expanding AI model sizes put extraordinary pressure on 1) AI Hardware Performance, 2) AI Model Efficiency and 3) AI Infrastructure Scaling. Companies are rapidly scaling up their AI models, betting on the premise that larger, more complex models will unlock new capabilities and opportunities, potentially leading to Artificial General Intelligence (AGI).

The Dominance of Global Giants: Developing these frontier models is not just a technical challenge but also a significant financial undertaking. Only a handful of global companies have the financial muscle and technical prowess to compete in this arena. This concentration of capability leads to a scenario where AI's future is shaped by a few major players, each vying for technological supremacy.

The Implications of Scale: The race for scale in AI development has far-reaching implications. It's not just about who can build the biggest model; it's about who can effectively harness such a model to solve complex, real-world problems. These will come from use cases built on top of many of these frontier platforms. This race will likely lead to breakthroughs in various fields, from healthcare to autonomous systems, fundamentally altering the technological landscape.

Navigating the Future

Strategic Partnerships and Collaborations: For companies outside the circle of these global giants, collaborations and building on top of their platforms is vital. Engaging with leaders in AI can provide access to frontier technologies, enabling smaller players to leverage these advancements for their own competitive advantage.

Investing in Complementary Technologies: Companies must also invest in complementary technologies and infrastructure that can support the integration and application of these frontier models. This includes cloud computing, data analytics, and cybersecurity.

---

Standalone AI Does Not Create Value

Despite considerable investments and high expectations, many companies find themselves grappling with the harsh reality that AI alone doesn't automatically translate into business success. A 2019 survey revealed that 40% of businesses failed to realize the anticipated benefits from their AI investments, highlighting a crucial lesson: Like the business IT revolution in the second quarter of the last century, AI is not a standalone solution for value creation but a tool that must be strategically integrated into broader business processes and objectives.

Understanding AI's Role in Business

The allure of AI has captivated many, yet its standalone implementation often falls short of delivering tangible business benefits. This disconnect highlights the need for a more nuanced understanding of AI's role in value creation.

The Necessity of Complementary Assets: Like any transformative innovation, AI requires a supportive ecosystem to flourish. This necessity extends beyond the technology itself to include compatible products, technologies, or services. Just as mobile phones require network connectivity and airlines need airports, AI needs an enabling environment to unlock its potential.

Embedding AI for Customer Value: The true power of AI emerges when it is intricately woven into complementary offerings. This integration ensures that AI not only functions efficiently but also enhances the overall value proposition to the customer.

The Limited Reach of Standalone AI: Companies lacking the capabilities or resources to offer these complementary elements may find their ability to leverage AI constrained. Without a holistic approach, AI's promise remains unfulfilled.

The crux of AI-driven value lies in the broader ecosystem - market access, hardware sensors, data, and technical expertise. Understanding and investing in these areas is crucial for businesses to fully leverage AI. Success with AI demands more than just adopting the technology; it requires a strategic vision to integrate AI. To move beyond the hype, businesses must critically assess how AI fits into their broader strategy and ecosystem, ensuring that it complements and enhances their core offerings.

IT is a Cost, AI Creates Value

In today's digitized world, the dilemma of 'Buy it or Build it' resonates strongly, especially when juxtaposing Information Technology (IT) and Artificial Intelligence (AI). While IT forms the backbone of modern enterprises, often categorized as a necessary yet undifferentiated operating cost, AI emerges as a transformative General-Purpose Technology (GPT) that unlocks value from data, enhancing product offerings and competitive edge.

The Strategic Crossroads: AI vs. IT

IT: Essential but Standardized: IT services, encompassing infrastructure like servers, networks, and software, are indispensable for business operations. However, they represent a baseline requirement, a cost of doing business that, while critical, offers limited differentiation. Many companies, irrespective of size, rely on external IT service providers for efficiency and cost-effectiveness, as the value lies in stable operations rather than innovation.

AI: A Catalyst for Value Creation: AI stands in stark contrast to traditional IT. It’s not just a tool but a driver of innovation, capable of analyzing vast datasets to reveal insights, optimize processes, and even predict market trends. AI's ability to transform data into actionable intelligence makes it a source of competitive advantage, rather than just a cost center.

Outsourcing AI vs. Outsourcing IT

Outsourcing IT – A Practical Necessity: For most businesses, outsourcing IT is a practical decision driven by cost, efficiency, and access to specialized expertise. It's a model that works because IT needs, while complex, are relatively standardized across industries.

Outsourcing AI – A Double-Edged Sword: The decision to outsource AI through a Data Analysis Intermediary (DAI) is more nuanced. On one hand, external AI providers offer advanced capabilities and expertise that may be too costly or complex to develop in-house. On the other hand, AI’s strategic value is closely tied to a company's unique data and specific operational needs. Outsourcing AI risks losing control over these critical strategic assets.

The Balance of Innovation and Control: The key in AI adoption is balancing innovation with strategic control. While outsourcing can accelerate AI integration and bring immediate benefits, building in-house AI capabilities can offer long-term strategic advantages, ensuring that the value derived from AI is unique and tailored to the company's specific needs.

The 'Buy it or Build it' question for AI in business strategy is not just about technical capability but about sustaining competitive advantage. While IT services can be efficiently outsourced as operational necessities, AI demands a more considered approach. For businesses looking to not just survive but thrive in the AI era, the focus should be on how to best leverage AI to extract unique value from their data and operations. Whether through outsourcing for speed and expertise or investing in in-house development for strategic control, the ultimate goal is to harness AI as a tool for innovation, growth, and market differentiation.

Buy it or Build it: Downsides of Outsourcing AI

1. Intensive data sharing needed: Maximizing AI benefits requires sharing sensitive information about products and assets, where the DAI lacks expertise.

2. Net zero competitive advantage: Using data, a DAI will enhance services for competitors, which offers no unique advantage to to the data owner.

3. Creates dependence on a black box:  A DAI’s influence over a companies systems will create a strategic dependency on opaque technology.

4. Data Science as a sole offering: Data owners generally already have the necessary market access, sector expertise and often data infrastructure. AI models are all open source and don’t offer a standalone competitive advantage. DAI’s in some industries will look more like undifferentiated data management and analysis services which use open source tools.

---

DAI Exception, Sensitive Industries

Data Analysis Intermediaries (DAIs) are emerging as vital players in sectors where direct data exchange is mired in conflicts of interest, such as the pharmaceutical industry. Their role as neutral intermediaries enables them to navigate sensitive data landscapes, particularly where direct transactions between data sources (like clinics) and end-users (like pharma companies) are problematic. However, while DAIs will become a ‘must have’ for companies in restricted industries to remain competitive, they will provide a 'net zero' competitive edge when servicing multiple clients within the same industry.

The Neutral Middleman in High-Stakes Industries

In industries like pharmaceuticals, direct data transactions between clinics and pharma companies can be fraught with ethical and legal complexities. DAIs step in as neutral parties, bridging this gap without the direct conflict of interest that might plague clinics or pharmaceutical companies. The company Prospection exemplifies this intermediary role by purchasing health data from clinics and turning it into actionable insights for pharmaceutical companies. This arrangement allows for data utilization without direct conflict, ensuring that sensitive health data is used ethically and responsibly.

The Value Proposition of DAIs

1. Expertise in Data Processing and Analysis: DAIs specialize in deciphering complex data sets, a task that requires a unique blend of technical skill and industry knowledge. This expertise allows them to extract meaningful patterns and insights that would be challenging for companies to achieve internally.

2. Customized Insights for Targeted Strategies: The insights provided by DAIs are highly customized. In the case of Prospection, the analysis can reveal specific market needs, effectiveness of drugs, and patient outcomes, enabling pharmaceutical companies to tailor their strategies effectively.

3. Cost-Effective Solution for Data-Driven Decisions: For many companies, particularly in specialized sectors like healthcare, developing in-house data analysis capabilities can be prohibitively expensive. DAIs offer a cost-effective alternative, providing high-quality insights without the need for significant internal investment.

DAIs play a pivotal role in the data ecosystem. They act as the crucial link between data collection and its practical application. By analyzing raw data, they transform it into strategic insights, thereby enabling companies to make informed decisions. Prospection stands out as a case study in the healthcare sector. By purchasing data related to drug usage and health condition markers, they apply sophisticated analytical methods to uncover patterns and trends. These insights are invaluable for pharmaceutical companies, helping them refine their marketing strategies and optimize drug development.

The Double-Edged Sword of DAI Services

1. Net Zero Competitive Edge: While DAIs offer valuable insights, their service to multiple clients within the same industry can lead to a 'net zero' competitive advantage. If all competitors have access to similar insights from the same DAI, the competitive edge that such insights might offer is essentially neutralized.

2. Dependence on External Insights: Reliance on DAIs for critical data analysis can lead companies to become overly dependent on external sources for strategic decision-making. This reliance can limit internal development of data capabilities and market differentiation.

Strategic Implications for Companies Using DAIs

1. Balancing External Insights with Internal Capabilities: Companies need to leverage the insights provided by DAIs while also investing in their internal data analysis capacities. This dual approach ensures diversity in strategic data use and reduces over-reliance on external sources.

2. Differentiating Through Customized Application: To overcome the 'net zero' competitive challenge, companies need to apply DAI-provided insights in unique ways, integrating them into broader, customized strategies that differentiate them from their competitors.

DAIs can offer a crucial service in complex markets like pharmaceuticals, acting as neutral intermediaries in scenarios where direct data exchange is not feasible or ethical. However, their role as providers to multiple industry players can equalize the competitive landscape, necessitating innovative and strategic application of their insights. Companies will need to balance the use of DAIs with the development of internal capabilities and ethical considerations, ensuring that they maximize the value of data while maintaining competitive differentiation and regulatory compliance.

---

Key Takeaways